Educational Process Mining¶

https://archive-beta.ics.uci.edu/dataset/346/educational+process+mining+epm+a+learning+analytics+data+set

Educational Process Mining (EPM): A Learning Analytics Data Set. (2015). UCI Machine Learning Repository.

This notebook includes exploratory data analysis of inputs, outputs, and inputs to outputs.

Inputs and Intermediate Output¶

There is an interaction between the continuous variables and the categorical variable actv_grp. As revealed by box plots, the distribution of interim scores is different depending on the value of actv_grp as follows:

  • Sessions 2 and 3: interim_score changes when actv_grp is 'FSM_Related' or 'Study Materials'
  • Sessions 4 and 5: interim_score changes when actv_grp is 'FSM_Related'
  • Session 6: interim_score changes when actv_grp is 'FSM' or 'Study Materials'

Scatter plots show that the relationships between interim_score and the continuous variables change depending on the value of actv_grp. For different values of actv_grp this relationship is stronger, weaker, or a different direction (positive/negative).

Inputs and Final Output¶

There is a positive relationship between the final score and final questions final_Es_4_2, final_Es_6_1, and final_Es_6_2. If students did well on these questions, they did well on the exam. Together these questions are worth 50% of the final score.

The total time spent did not appear to have an effect on whether a student passed the final. However, the number of sessions and the variety of exercises and activities in which students participated did have an effect. While participating in more sessions and in a variety of exercise and activities did not guarantee a passing score, if students did not participate in at least four sessions (with the exception of two students) and in a variety of exercises and activities, they did not pass.

In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

import seaborn as sns

Load the data from github repository¶

In [2]:
inputs_path_1 = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/inputs_clean_p1.csv'
inputs_path_2 = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/inputs_clean_p2.csv'
inputs_path_3 = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/inputs_clean_p3.csv'
interim_path = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/interim_df.csv'
final_path = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/final_df.csv'
pts_path = 'https://raw.githubusercontent.com/lisaover/CMPINF2120_project/main/final_points_lookup.csv'
In [3]:
inputs_p1 = pd.read_csv(inputs_path_1)
In [4]:
inputs_p2 = pd.read_csv(inputs_path_2)
In [5]:
inputs_p3 = pd.read_csv(inputs_path_3)
In [6]:
inputs_init = pd.concat([inputs_p1, inputs_p2, inputs_p3], ignore_index=True, axis=0)
In [7]:
interim_init = pd.read_csv(interim_path)
In [8]:
final_init = pd.read_csv(final_path)
In [9]:
pts_final_lookup = pd.read_csv(pts_path)
In [10]:
inputs_init.columns
Out[10]:
Index(['sess', 'sid', 'start_dt', 'end_dt', 'exer', 'actv', 'actv_grp',
       'total_ms', 'idle_ms', 'mw', 'mwc', 'mcl', 'mcr', 'mm', 'ks',
       'total_ms_wdz', 'idle_ms_wdz', 'mw_wdz', 'mwc_wdz', 'mcl_wdz',
       'mcr_wdz', 'mm_wdz', 'ks_wdz', 'total_ms_sqrt', 'idle_ms_sqrt',
       'mw_sqrt', 'mwc_sqrt', 'mcl_sqrt', 'mcr_sqrt', 'mm_sqrt', 'ks_sqrt'],
      dtype='object')
In [11]:
interim_init.columns
Out[11]:
Index(['sid', 'interim_scr2', 'interim_scr3', 'interim_scr4', 'interim_scr5',
       'interim_scr6'],
      dtype='object')
In [12]:
final_init.columns
Out[12]:
Index(['sid', 'Es_1q1', 'Es_1q2', 'Es_2q1', 'Es_2q2', 'Es_3q1', 'Es_3q2',
       'Es_3q3', 'Es_3q4', 'Es_3q5', 'Es_4q1', 'Es_4q2', 'Es_5q1', 'Es_5q2',
       'Es_5q3', 'Es_6q1', 'Es_6q2', 'final_score'],
      dtype='object')
In [13]:
pts_final_lookup.columns
Out[13]:
Index(['question', 'question_points'], dtype='object')

Inputs¶

In [14]:
inputs_init.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 227129 entries, 0 to 227128
Data columns (total 31 columns):
 #   Column         Non-Null Count   Dtype  
---  ------         --------------   -----  
 0   sess           227129 non-null  int64  
 1   sid            227129 non-null  int64  
 2   start_dt       227129 non-null  object 
 3   end_dt         227128 non-null  object 
 4   exer           227129 non-null  object 
 5   actv           227129 non-null  object 
 6   actv_grp       227129 non-null  object 
 7   total_ms       227128 non-null  float64
 8   idle_ms        226971 non-null  float64
 9   mw             227129 non-null  int64  
 10  mwc            227129 non-null  int64  
 11  mcl            227129 non-null  int64  
 12  mcr            227129 non-null  int64  
 13  mm             227129 non-null  int64  
 14  ks             227129 non-null  int64  
 15  total_ms_wdz   227129 non-null  float64
 16  idle_ms_wdz    227129 non-null  float64
 17  mw_wdz         227129 non-null  float64
 18  mwc_wdz        227129 non-null  float64
 19  mcl_wdz        227129 non-null  float64
 20  mcr_wdz        227129 non-null  float64
 21  mm_wdz         227129 non-null  float64
 22  ks_wdz         227129 non-null  float64
 23  total_ms_sqrt  227128 non-null  float64
 24  idle_ms_sqrt   226971 non-null  float64
 25  mw_sqrt        227129 non-null  float64
 26  mwc_sqrt       227129 non-null  float64
 27  mcl_sqrt       227129 non-null  float64
 28  mcr_sqrt       227129 non-null  float64
 29  mm_sqrt        227129 non-null  float64
 30  ks_sqrt        227129 non-null  float64
dtypes: float64(18), int64(8), object(5)
memory usage: 53.7+ MB
In [15]:
inputs_init.isna().sum()
Out[15]:
sess               0
sid                0
start_dt           0
end_dt             1
exer               0
actv               0
actv_grp           0
total_ms           1
idle_ms          158
mw                 0
mwc                0
mcl                0
mcr                0
mm                 0
ks                 0
total_ms_wdz       0
idle_ms_wdz        0
mw_wdz             0
mwc_wdz            0
mcl_wdz            0
mcr_wdz            0
mm_wdz             0
ks_wdz             0
total_ms_sqrt      1
idle_ms_sqrt     158
mw_sqrt            0
mwc_sqrt           0
mcl_sqrt           0
mcr_sqrt           0
mm_sqrt            0
ks_sqrt            0
dtype: int64
In [16]:
inputs_init['sid'] = inputs_init['sid'].astype('object')
inputs_init['sess'] = inputs_init['sess'].astype('object')
In [17]:
def get_var_list(df, lst):
    temp = []
    for s in lst:
        temp = [j for i in [temp, df.loc[:, df.columns.str.endswith(s)].columns] for j in i] 
    return temp
In [18]:
sqrt_vars = get_var_list(inputs_init,['sqrt'])
wdz_vars = get_var_list(inputs_init,['wdz'])
sqrt_ms_vars = get_var_list(inputs_init,['ms_sqrt'])
wdz_ms_vars = get_var_list(inputs_init,['ms_wdz'])
orig_ms_vars = get_var_list(inputs_init, ['ms'])
cat_vars = inputs_init.drop(columns=['actv','start_dt','end_dt']).copy().select_dtypes('object').columns.to_list()
In [19]:
num_vars = inputs_init.select_dtypes('number').columns.to_list()
num_vars
Out[19]:
['total_ms',
 'idle_ms',
 'mw',
 'mwc',
 'mcl',
 'mcr',
 'mm',
 'ks',
 'total_ms_wdz',
 'idle_ms_wdz',
 'mw_wdz',
 'mwc_wdz',
 'mcl_wdz',
 'mcr_wdz',
 'mm_wdz',
 'ks_wdz',
 'total_ms_sqrt',
 'idle_ms_sqrt',
 'mw_sqrt',
 'mwc_sqrt',
 'mcl_sqrt',
 'mcr_sqrt',
 'mm_sqrt',
 'ks_sqrt']
In [20]:
orig_vars = [v for v in num_vars if v not in sqrt_vars + wdz_vars]
orig_vars
Out[20]:
['total_ms', 'idle_ms', 'mw', 'mwc', 'mcl', 'mcr', 'mm', 'ks']
In [21]:
activity_orig_vars = [v for v in orig_vars if v not in orig_ms_vars]
activity_orig_vars
Out[21]:
['mw', 'mwc', 'mcl', 'mcr', 'mm', 'ks']
In [22]:
activity_sqrt_vars = [v for v in sqrt_vars if v not in sqrt_ms_vars]
activity_sqrt_vars
Out[22]:
['mw_sqrt', 'mwc_sqrt', 'mcl_sqrt', 'mcr_sqrt', 'mm_sqrt', 'ks_sqrt']
In [23]:
activity_wdz_vars = [v for v in wdz_vars if v not in wdz_ms_vars]
activity_wdz_vars
Out[23]:
['mw_wdz', 'mwc_wdz', 'mcl_wdz', 'mcr_wdz', 'mm_wdz', 'ks_wdz']
In [24]:
cat_vars
Out[24]:
['sess', 'sid', 'exer', 'actv_grp']
In [25]:
sqrt_vars
Out[25]:
['total_ms_sqrt',
 'idle_ms_sqrt',
 'mw_sqrt',
 'mwc_sqrt',
 'mcl_sqrt',
 'mcr_sqrt',
 'mm_sqrt',
 'ks_sqrt']
In [26]:
wdz_vars
Out[26]:
['total_ms_wdz',
 'idle_ms_wdz',
 'mw_wdz',
 'mwc_wdz',
 'mcl_wdz',
 'mcr_wdz',
 'mm_wdz',
 'ks_wdz']

Outputs¶

In [27]:
interim_init.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 115 entries, 0 to 114
Data columns (total 6 columns):
 #   Column        Non-Null Count  Dtype  
---  ------        --------------  -----  
 0   sid           115 non-null    int64  
 1   interim_scr2  115 non-null    float64
 2   interim_scr3  115 non-null    float64
 3   interim_scr4  115 non-null    float64
 4   interim_scr5  115 non-null    float64
 5   interim_scr6  115 non-null    float64
dtypes: float64(5), int64(1)
memory usage: 5.5 KB
In [28]:
interim_init.isna().sum()
Out[28]:
sid             0
interim_scr2    0
interim_scr3    0
interim_scr4    0
interim_scr5    0
interim_scr6    0
dtype: int64
In [29]:
final_init.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 62 entries, 0 to 61
Data columns (total 18 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   sid          62 non-null     int64  
 1   Es_1q1       62 non-null     float64
 2   Es_1q2       62 non-null     float64
 3   Es_2q1       62 non-null     float64
 4   Es_2q2       62 non-null     float64
 5   Es_3q1       62 non-null     float64
 6   Es_3q2       62 non-null     float64
 7   Es_3q3       62 non-null     float64
 8   Es_3q4       62 non-null     float64
 9   Es_3q5       62 non-null     float64
 10  Es_4q1       62 non-null     float64
 11  Es_4q2       62 non-null     float64
 12  Es_5q1       62 non-null     float64
 13  Es_5q2       62 non-null     float64
 14  Es_5q3       62 non-null     float64
 15  Es_6q1       62 non-null     float64
 16  Es_6q2       62 non-null     float64
 17  final_score  62 non-null     float64
dtypes: float64(17), int64(1)
memory usage: 8.8 KB
In [30]:
final_init.isna().sum()
Out[30]:
sid            0
Es_1q1         0
Es_1q2         0
Es_2q1         0
Es_2q2         0
Es_3q1         0
Es_3q2         0
Es_3q3         0
Es_3q4         0
Es_3q5         0
Es_4q1         0
Es_4q2         0
Es_5q1         0
Es_5q2         0
Es_5q3         0
Es_6q1         0
Es_6q2         0
final_score    0
dtype: int64
In [31]:
pts_final_lookup.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 17 entries, 0 to 16
Data columns (total 2 columns):
 #   Column           Non-Null Count  Dtype 
---  ------           --------------  ----- 
 0   question         17 non-null     object
 1   question_points  17 non-null     int64 
dtypes: int64(1), object(1)
memory usage: 400.0+ bytes
In [32]:
pts_final_lookup.question.unique()
Out[32]:
array(['Es_1q1', 'Es_1q2', 'Es_2q1', 'Es_2q2', 'Es_3q1', 'Es_3q2',
       'Es_3q3', 'Es_3q4', 'Es_3q5', 'Es_4q1', 'Es_4q2', 'Es_5q1',
       'Es_5q2', 'Es_5q3', 'Es_6q1', 'Es_6q2', 'final_score'],
      dtype=object)

Add classification variable to indicate event 'pass' if composite score is greater than or equal to 70¶

In [33]:
final_init['pass'] = np.where( final_init['final_score'] >= 70, 1, 0 )

Explore the input data¶

In [34]:
inputs_df = inputs_init.copy()
In [35]:
inputs_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 227129 entries, 0 to 227128
Data columns (total 31 columns):
 #   Column         Non-Null Count   Dtype  
---  ------         --------------   -----  
 0   sess           227129 non-null  object 
 1   sid            227129 non-null  object 
 2   start_dt       227129 non-null  object 
 3   end_dt         227128 non-null  object 
 4   exer           227129 non-null  object 
 5   actv           227129 non-null  object 
 6   actv_grp       227129 non-null  object 
 7   total_ms       227128 non-null  float64
 8   idle_ms        226971 non-null  float64
 9   mw             227129 non-null  int64  
 10  mwc            227129 non-null  int64  
 11  mcl            227129 non-null  int64  
 12  mcr            227129 non-null  int64  
 13  mm             227129 non-null  int64  
 14  ks             227129 non-null  int64  
 15  total_ms_wdz   227129 non-null  float64
 16  idle_ms_wdz    227129 non-null  float64
 17  mw_wdz         227129 non-null  float64
 18  mwc_wdz        227129 non-null  float64
 19  mcl_wdz        227129 non-null  float64
 20  mcr_wdz        227129 non-null  float64
 21  mm_wdz         227129 non-null  float64
 22  ks_wdz         227129 non-null  float64
 23  total_ms_sqrt  227128 non-null  float64
 24  idle_ms_sqrt   226971 non-null  float64
 25  mw_sqrt        227129 non-null  float64
 26  mwc_sqrt       227129 non-null  float64
 27  mcl_sqrt       227129 non-null  float64
 28  mcr_sqrt       227129 non-null  float64
 29  mm_sqrt        227129 non-null  float64
 30  ks_sqrt        227129 non-null  float64
dtypes: float64(18), int64(6), object(7)
memory usage: 53.7+ MB
In [36]:
inputs_df.exer.value_counts()
Out[36]:
Es_4_5    22091
Es_6_3    19290
Es_6_2    15996
Es_6_1    13505
Es_5_4    12217
Es_1_1    11406
Es_2_2    10414
Es_3_2    10131
Es_1_2     9614
Es_5_1     9090
Es_2_1     7952
Es_3_1     7637
Es_2_4     7559
Es_5_3     7219
Es_1_3     6716
Es_2_3     6082
Es_4_1     5727
Es_1_4     5502
Es_3_4     5344
Es_5_2     5156
Es_3_3     4680
Es_2_5     4241
Es_4_2     4044
Es_4_3     3757
Es_4_4     3364
Es_6_4     2939
Es         2647
Es_2_6     2573
Es_6_6      135
Es_6_5      101
Name: exer, dtype: int64

Aggregate by student to count the unique values of sess, exer, actv, and actv_grp for each student¶

In [37]:
sid_counts = inputs_df.groupby(['sid']).\
            aggregate(sess_uniq = ('sess', 'nunique'),
                      exer_uniq = ('exer', 'nunique'),
                      actv_uniq = ('actv', 'nunique'),
                      actv_grp_uniq = ('actv_grp', 'nunique')).\
            reset_index().copy()
In [38]:
sid_counts.head()
Out[38]:
sid sess_uniq exer_uniq actv_uniq actv_grp_uniq
0 1 5 23 78 10
1 2 6 30 92 10
2 3 4 19 61 8
3 4 6 27 83 10
4 5 6 27 83 10
In [39]:
sid_counts.loc[sid_counts['actv_uniq'] == max(sid_counts['actv_uniq'])]
Out[39]:
sid sess_uniq exer_uniq actv_uniq actv_grp_uniq
31 32 6 30 93 11
43 44 6 29 93 11
51 52 6 29 93 10
69 70 6 30 93 10
80 81 6 30 93 10
84 85 6 29 93 11
93 94 6 30 93 10

Bar charts show progression through the course by participation in exercises and activities for student 32 who has high partitipation.¶

Each record represents an interval of time during which a student performed or participated in an activity, related to a particular exercise, within a session.¶

The actv variable value is a composite value consisting of an activity type and an exercise. The activity type was extracted from actv to create a new actv_grp.

The bar chart below illustrates how student 32 progressses through the exercises (exer) and activity types (actv_grp) for each session (sess).

In [40]:
sns.set_style('whitegrid')

sns.catplot(data = inputs_df.loc[inputs_df['sid'] == 32], x='exer', col='sess', hue='actv_grp', 
            col_wrap=1, kind='count', palette='deep',
            height=3, aspect=6, sharey=False, sharex=False)

plt.show()

The bar chart below illustrates how student 32 progressses through the sessions (sess) by activity type (actv_grp).

In [41]:
sns.set_style('whitegrid')

sns.catplot(data = inputs_df.loc[inputs_df['sid'] == 32], x='exer', col='actv_grp', hue='sess', 
            col_wrap=1, kind='count', palette='deep',
            height=3, aspect=6, sharey=False, sharex=False)

plt.show()
In [42]:
inputs_df.loc[inputs_df['sid'].isin([32, 44, 85])]
Out[42]:
sess sid start_dt end_dt exer actv actv_grp total_ms idle_ms mw ... mm_wdz ks_wdz total_ms_sqrt idle_ms_sqrt mw_sqrt mwc_sqrt mcl_sqrt mcr_sqrt mm_sqrt ks_sqrt
7807 1 32 2014-02-10 11:24:18 2014-02-10 11:24:22 Es Other Other 4000.0 4242.0 0 ... 79.0 0.0 63.245553 65.130638 0.0 0.0 0.000000 0.0 8.888194 0.0
7808 1 32 2014-02-10 11:24:23 2014-02-10 11:24:23 Es Other Other 0.0 0.0 0 ... 43.0 0.0 0.000000 0.000000 0.0 0.0 1.414214 0.0 6.557439 0.0
7809 1 32 2014-02-10 11:24:24 2014-02-10 11:24:24 Es Other Other 0.0 0.0 0 ... 58.0 0.0 0.000000 0.000000 0.0 0.0 1.414214 0.0 7.615773 0.0
7810 1 32 2014-02-10 11:24:25 2014-02-10 11:24:46 Es Other Other 21000.0 185772.0 0 ... 82.0 0.0 144.913767 431.012761 0.0 0.0 1.414214 0.0 9.055385 0.0
7811 1 32 2014-02-10 11:24:47 2014-02-10 11:24:49 Es Other Other 2000.0 94.0 0 ... 194.0 0.0 44.721360 9.695360 0.0 0.0 1.414214 0.0 13.928388 0.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
218288 6 85 2014-11-12 13:55:44 2014-11-12 13:55:45 Es_6_6 Other Other 1000.0 124.0 0 ... 90.0 0.0 31.622777 11.135529 0.0 0.0 1.414214 0.0 9.486833 0.0
218289 6 85 2014-11-12 13:55:46 2014-11-12 13:55:58 Es_6_6 Study_Es_6_6 Study 12000.0 19722.0 0 ... 217.0 0.0 109.544512 140.435038 0.0 0.0 2.000000 0.0 14.730920 0.0
218290 6 85 2014-11-12 13:55:59 2014-11-12 13:55:59 Es_6_6 Blank Blank 0.0 0.0 0 ... 61.0 0.0 0.000000 0.000000 0.0 0.0 1.414214 0.0 7.810250 0.0
218291 6 85 2014-11-12 13:56:00 2014-11-12 13:56:04 Es_6_6 Aulaweb Aulaweb 4000.0 0.0 0 ... 193.0 0.0 63.245553 0.000000 0.0 0.0 2.000000 0.0 13.892444 0.0
218292 6 85 2014-11-12 13:56:05 2014-11-12 13:56:08 Es_6_6 Aulaweb Aulaweb 3000.0 0.0 0 ... 98.0 0.0 54.772256 0.000000 0.0 0.0 0.000000 0.0 9.899495 0.0

9806 rows × 31 columns

The bar chart below illustrates how students 32, 44, and 85 progress through the activities (actv), which is the combined exercise and activity type, by session.

In [43]:
sns.set_style('whitegrid')

sns.catplot(data = inputs_df.loc[inputs_df['sid'].isin([32, 44, 85])], x='actv', col='sess', hue='sid', 
            col_wrap=1, kind='count', palette='deep',
            height=5, aspect=10, sharey=False, sharex=False)

plt.show()

Select 3 students with high participation and 3 students with low participation for heatmaps¶

In [44]:
sid_counts.loc[sid_counts['actv_uniq'] == max(sid_counts['actv_uniq'])]
Out[44]:
sid sess_uniq exer_uniq actv_uniq actv_grp_uniq
31 32 6 30 93 11
43 44 6 29 93 11
51 52 6 29 93 10
69 70 6 30 93 10
80 81 6 30 93 10
84 85 6 29 93 11
93 94 6 30 93 10
In [45]:
sid_counts.loc[sid_counts['actv_uniq'] <= np.quantile(sid_counts['actv_uniq'], 0.15)]
Out[45]:
sid sess_uniq exer_uniq actv_uniq actv_grp_uniq
25 26 1 6 9 5
32 33 1 6 20 8
56 57 2 10 30 10
57 58 1 6 16 6
61 62 2 10 34 8
76 77 1 6 16 6
88 89 4 12 28 9
102 103 2 10 34 9
104 105 2 11 26 8
105 106 1 5 19 9
106 108 1 5 22 9
107 109 1 5 20 8
108 110 1 5 18 8
109 111 1 5 20 9
110 112 1 5 19 8
111 113 1 5 19 8
112 114 1 5 19 8
In [46]:
sid_counts.loc[sid_counts['actv_uniq'] == max(sid_counts['actv_uniq'])]
Out[46]:
sid sess_uniq exer_uniq actv_uniq actv_grp_uniq
31 32 6 30 93 11
43 44 6 29 93 11
51 52 6 29 93 10
69 70 6 30 93 10
80 81 6 30 93 10
84 85 6 29 93 11
93 94 6 30 93 10

Aggregate by student and session to count the number of unique values of exer, actv, and actv_grp for each student and session¶

In [47]:
sid_sess_counts = inputs_df.groupby(['sid', 'sess']).\
            aggregate(exer_uniq = ('exer', 'nunique'),
                      actv_uniq = ('actv', 'nunique'),
                      actv_grp_uniq = ('actv_grp', 'nunique')).\
            reset_index().copy()
In [48]:
sid_sess_counts
Out[48]:
sid sess exer_uniq actv_uniq actv_grp_uniq
0 1 1 5 20 8
1 1 2 7 25 8
2 1 4 6 23 9
3 1 5 5 18 8
4 1 6 4 18 10
... ... ... ... ... ...
509 110 1 5 18 8
510 111 1 5 20 9
511 112 1 5 19 8
512 113 1 5 19 8
513 114 1 5 19 8

514 rows × 5 columns

In [49]:
sid_sess_counts.loc[ (sid_sess_counts.sid.isin([32, 44, 85]))]
Out[49]:
sid sess exer_uniq actv_uniq actv_grp_uniq
152 32 1 5 19 8
153 32 2 7 25 9
154 32 3 5 18 8
155 32 4 6 21 8
156 32 5 5 18 8
157 32 6 7 23 10
208 44 1 5 18 8
209 44 2 7 26 9
210 44 3 5 19 9
211 44 4 6 21 8
212 44 5 5 18 9
213 44 6 6 24 11
404 85 1 5 18 8
405 85 2 7 24 8
406 85 3 5 19 9
407 85 4 6 21 8
408 85 5 6 22 8
409 85 6 6 23 10
In [50]:
sid_sess_counts.loc[ (sid_sess_counts.sid.isin([57, 77, 89]))]
Out[50]:
sid sess exer_uniq actv_uniq actv_grp_uniq
277 57 4 6 16 6
278 57 6 5 20 10
365 77 4 6 16 6
428 89 3 4 11 7
429 89 4 6 17 7
430 89 5 2 6 5
431 89 6 3 10 8
In [51]:
uniq_vars = get_var_list(sid_sess_counts,['uniq'])
In [52]:
sid_3high_sess_counts = sid_sess_counts.loc[sid_sess_counts.sid.isin([32, 44, 85])].copy()
In [53]:
sid_3high_sess_counts.head()
Out[53]:
sid sess exer_uniq actv_uniq actv_grp_uniq
152 32 1 5 19 8
153 32 2 7 25 9
154 32 3 5 18 8
155 32 4 6 21 8
156 32 5 5 18 8
In [54]:
the_sessions = sid_3high_sess_counts.sess.unique().tolist()
In [55]:
sid_3high_sess_counts.reset_index()[['sid','sess', uniq_vars[0]]].pivot(index='sid', columns='sess', values=uniq_vars[0])
Out[55]:
sess 1 2 3 4 5 6
sid
32 5 7 5 6 5 7
44 5 7 5 6 5 6
85 5 7 5 6 6 6

Heatmaps show participation as the unique count of exercises, activities, and activity groups for the selected students with high and low participation¶

In [56]:
fig, axs = plt.subplots(1, len(uniq_vars), figsize=(18,6), sharex=True, sharey=True)

for ix, x in enumerate(uniq_vars):
    sns.heatmap(data = sid_3high_sess_counts.reset_index()[['sid','sess', x]].pivot(index='sid', columns='sess', values=x),
                center=sid_3high_sess_counts.reset_index()[x].mean(),
                cmap='coolwarm', cbar=False,
                annot=True, annot_kws={'size': 25},
                ax=axs[ix])
    
    axs[ix].set_title(str(x))

plt.show()
In [57]:
sid_sess_counts.loc[ (sid_sess_counts.sid.isin([57, 77, 89]))]
Out[57]:
sid sess exer_uniq actv_uniq actv_grp_uniq
277 57 4 6 16 6
278 57 6 5 20 10
365 77 4 6 16 6
428 89 3 4 11 7
429 89 4 6 17 7
430 89 5 2 6 5
431 89 6 3 10 8
In [58]:
sid_3low_sess_counts = sid_sess_counts.loc[sid_sess_counts.sid.isin([57, 77, 89])].copy()
In [59]:
the_sessions_low = sid_3low_sess_counts.sess.unique().tolist()
In [60]:
fig, axs = plt.subplots(1, len(uniq_vars), figsize=(18,6), sharex=True, sharey=True)

for ix, x in enumerate(uniq_vars):
    sns.heatmap(data = sid_3low_sess_counts.reset_index()[['sid','sess', x]].pivot(index='sid', columns='sess', values=x),
                center=sid_3low_sess_counts.reset_index()[x].mean(),
                cmap='coolwarm', cbar=False,
                annot=True, annot_kws={'size': 25},
                ax=axs[ix])
    
    axs[ix].set_title(str(x))

plt.show()

Scatter plots show unique exercises, activities, and activity groups counts by student and session for all students¶

In [61]:
sns.relplot(data = sid_sess_counts, x='sess', y='sid', hue='exer_uniq', height=10, aspect=2)

plt.show()
In [62]:
sns.relplot(data = sid_sess_counts, x='sess', y='sid', hue='actv_uniq', height=10, aspect=2)

plt.show()
In [63]:
sns.relplot(data = sid_sess_counts, x='sess', y='sid', hue='actv_grp_uniq', height=10, aspect=2)

plt.show()

The number of records per exercise by student¶

In [64]:
sns.set_style('whitegrid')

sns.catplot(data = inputs_df, x='exer', col='sid', col_wrap=1, kind='count', color='slategray',
            height=3, aspect=6, sharey=False, sharex=False)

plt.show()

The number of records per activity by student¶

In [65]:
inputs_df.actv_grp.value_counts()
Out[65]:
TextEditor         41900
Deeds              38000
Other              33009
Blank              23929
Study              21540
Diagram            20546
Properties         19419
FSM                17000
Aulaweb             8113
FSM_Related         3275
Study_Materials      398
Name: actv_grp, dtype: int64
In [66]:
sns.set_style('whitegrid')

sns.catplot(data = inputs_df, x='actv_grp', col='sid', col_wrap=1, kind='count', color='slategray',
            height=3, aspect=4, sharey=False, sharex=False)

plt.show()

The number of records per student by session¶

In [67]:
sns.catplot(data = inputs_df, x='sid', col='sess', col_wrap=1, kind='count', color='slategray',
            height=5, aspect=8, sharey=False)

plt.show()

The number of records per session by student¶

In [68]:
sns.catplot(data = inputs_df, x='sess', col='sid', col_wrap=2, kind='count', color='slategray',
            sharey=False, sharex=False)

plt.show()

The distribution of numeric variables (sqrt)¶

In [69]:
inputs_sqrt_df = inputs_df[cat_vars + sqrt_vars].copy()
In [70]:
inputs_sqrt_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 227129 entries, 0 to 227128
Data columns (total 12 columns):
 #   Column         Non-Null Count   Dtype  
---  ------         --------------   -----  
 0   sess           227129 non-null  object 
 1   sid            227129 non-null  object 
 2   exer           227129 non-null  object 
 3   actv_grp       227129 non-null  object 
 4   total_ms_sqrt  227128 non-null  float64
 5   idle_ms_sqrt   226971 non-null  float64
 6   mw_sqrt        227129 non-null  float64
 7   mwc_sqrt       227129 non-null  float64
 8   mcl_sqrt       227129 non-null  float64
 9   mcr_sqrt       227129 non-null  float64
 10  mm_sqrt        227129 non-null  float64
 11  ks_sqrt        227129 non-null  float64
dtypes: float64(8), object(4)
memory usage: 20.8+ MB
In [71]:
inputs_sqrt_df.describe()
Out[71]:
total_ms_sqrt idle_ms_sqrt mw_sqrt mwc_sqrt mcl_sqrt mcr_sqrt mm_sqrt ks_sqrt
count 227128.000000 226971.000000 227129.000000 227129.000000 227129.000000 227129.000000 227129.000000 227129.000000
mean 75.724962 221.136780 0.380204 0.002757 1.994965 0.114958 15.115336 0.661013
std 100.630429 1046.482145 1.600802 0.073777 1.759565 0.568873 13.643873 2.417640
min 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% 0.000000 0.000000 0.000000 0.000000 1.414214 0.000000 7.874008 0.000000
50% 44.721360 8.944272 0.000000 0.000000 1.414214 0.000000 11.704700 0.000000
75% 94.868330 75.766087 0.000000 0.000000 2.236068 0.000000 18.303005 0.000000
max 1964.942747 85116.014486 53.888774 7.745967 33.105891 12.961481 293.170599 68.949257
In [72]:
cat_vars
Out[72]:
['sess', 'sid', 'exer', 'actv_grp']
In [73]:
inputs_sqrt_lf = inputs_sqrt_df.melt( id_vars=cat_vars, 
                  value_vars=sqrt_vars).copy()
In [74]:
inputs_sqrt_lf.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1817032 entries, 0 to 1817031
Data columns (total 6 columns):
 #   Column    Dtype  
---  ------    -----  
 0   sess      object 
 1   sid       object 
 2   exer      object 
 3   actv_grp  object 
 4   variable  object 
 5   value     float64
dtypes: float64(1), object(5)
memory usage: 83.2+ MB
In [75]:
sns.catplot(data=inputs_sqrt_lf, x = 'variable', y = 'value', 
            col = 'sess', col_wrap=2, kind = 'box', 
            hue = 'actv_grp', height=3, aspect=4, 
            sharey=False, sharex=False)

plt.show()
In [76]:
sns.catplot(data=inputs_sqrt_lf, x = 'variable', y = 'value', 
            col = 'actv_grp', col_wrap=2, kind = 'box', 
            hue = 'actv_grp', height=3, aspect=4, 
            sharey=False, sharex=False)

plt.show()
In [77]:
sns.catplot(data = inputs_sqrt_df, x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [78]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==1], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [79]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==2], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [80]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==3], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [81]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==4], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [82]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==5], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()
In [83]:
sns.catplot(data = inputs_sqrt_df.loc[inputs_sqrt_df.sess==6], x='actv_grp', y='total_ms_sqrt', kind='box', aspect=3)

plt.show()

Correlation¶

In [84]:
inputs_sqrt_df.loc[:, sqrt_vars].corr()
Out[84]:
total_ms_sqrt idle_ms_sqrt mw_sqrt mwc_sqrt mcl_sqrt mcr_sqrt mm_sqrt ks_sqrt
total_ms_sqrt 1.000000 0.772576 0.349806 0.032966 0.674365 0.308610 0.738497 0.520151
idle_ms_sqrt 0.772576 1.000000 0.282011 0.023932 0.439299 0.166843 0.450366 0.442888
mw_sqrt 0.349806 0.282011 1.000000 0.050244 0.212130 0.040273 0.332769 0.302601
mwc_sqrt 0.032966 0.023932 0.050244 1.000000 0.007407 0.013707 0.038157 0.003945
mcl_sqrt 0.674365 0.439299 0.212130 0.007407 1.000000 0.432783 0.803341 0.358556
mcr_sqrt 0.308610 0.166843 0.040273 0.013707 0.432783 1.000000 0.390619 0.075800
mm_sqrt 0.738497 0.450366 0.332769 0.038157 0.803341 0.390619 1.000000 0.364175
ks_sqrt 0.520151 0.442888 0.302601 0.003945 0.358556 0.075800 0.364175 1.000000
In [85]:
fig, ax = plt.subplots(figsize=(12, 8))

sns.heatmap(data = inputs_sqrt_df.loc[:, sqrt_vars].corr(),
            vmin=-1, vmax=1, center=0,
            cmap='coolwarm',
            ax=ax)

plt.show()

Explore the intermediate output data¶

In [86]:
interim_df = interim_init.copy()
In [87]:
interim_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 115 entries, 0 to 114
Data columns (total 6 columns):
 #   Column        Non-Null Count  Dtype  
---  ------        --------------  -----  
 0   sid           115 non-null    int64  
 1   interim_scr2  115 non-null    float64
 2   interim_scr3  115 non-null    float64
 3   interim_scr4  115 non-null    float64
 4   interim_scr5  115 non-null    float64
 5   interim_scr6  115 non-null    float64
dtypes: float64(5), int64(1)
memory usage: 5.5 KB
In [88]:
interim_df
Out[88]:
sid interim_scr2 interim_scr3 interim_scr4 interim_scr5 interim_scr6
0 1 5.0 0.0 4.5 4.0 2.25
1 2 4.0 3.5 4.5 4.0 1.00
2 3 3.5 3.5 4.5 4.0 0.00
3 4 6.0 4.0 5.0 3.5 2.75
4 5 5.0 4.0 5.0 4.0 2.75
... ... ... ... ... ... ...
110 111 3.0 0.0 0.0 3.0 0.00
111 112 0.0 0.0 0.0 0.0 0.00
112 113 0.0 0.0 0.0 0.0 0.00
113 114 0.0 0.0 0.0 0.0 0.00
114 115 0.0 0.0 0.0 0.0 0.00

115 rows × 6 columns

In [89]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=interim_df, x='interim_scr2', ax=ax)

plt.show()
In [90]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=interim_df, x='interim_scr3', ax=ax)

plt.show()
In [91]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=interim_df, x='interim_scr4', ax=ax)

plt.show()
In [92]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=interim_df, x='interim_scr5', ax=ax)

plt.show()
In [93]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=interim_df, x='interim_scr6', ax=ax)

plt.show()

Point plots show the inermediate progress for each student¶

In [94]:
interim_lf = interim_df.melt( id_vars=['sid']).\
             rename(columns={"variable": "interim_session", "value": "interim_score"}).\
             copy()
In [95]:
interim_lf['sess'] = interim_lf.interim_session.str.extract('(\d+)')
In [96]:
interim_lf['sess'] = [x.strip() for x in interim_lf.sess]
In [97]:
interim_lf['sess'] = interim_lf['sess'].astype('int')
In [98]:
interim_lf
Out[98]:
sid interim_session interim_score sess
0 1 interim_scr2 5.0 2
1 2 interim_scr2 4.0 2
2 3 interim_scr2 3.5 2
3 4 interim_scr2 6.0 2
4 5 interim_scr2 5.0 2
... ... ... ... ...
570 111 interim_scr6 0.0 6
571 112 interim_scr6 0.0 6
572 113 interim_scr6 0.0 6
573 114 interim_scr6 0.0 6
574 115 interim_scr6 0.0 6

575 rows × 4 columns

In [99]:
interim_lf.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 575 entries, 0 to 574
Data columns (total 4 columns):
 #   Column           Non-Null Count  Dtype  
---  ------           --------------  -----  
 0   sid              575 non-null    int64  
 1   interim_session  575 non-null    object 
 2   interim_score    575 non-null    float64
 3   sess             575 non-null    int64  
dtypes: float64(1), int64(2), object(1)
memory usage: 18.1+ KB
In [100]:
sns.catplot(data=interim_lf, x='sess', y='interim_score', 
            col = 'sid', kind='point', color='slategray',
            col_wrap=2, sharey=False, sharex=False)

plt.show()

Explore the final output data¶

In [101]:
final_df = final_init.copy()
In [102]:
final_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 62 entries, 0 to 61
Data columns (total 19 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   sid          62 non-null     int64  
 1   Es_1q1       62 non-null     float64
 2   Es_1q2       62 non-null     float64
 3   Es_2q1       62 non-null     float64
 4   Es_2q2       62 non-null     float64
 5   Es_3q1       62 non-null     float64
 6   Es_3q2       62 non-null     float64
 7   Es_3q3       62 non-null     float64
 8   Es_3q4       62 non-null     float64
 9   Es_3q5       62 non-null     float64
 10  Es_4q1       62 non-null     float64
 11  Es_4q2       62 non-null     float64
 12  Es_5q1       62 non-null     float64
 13  Es_5q2       62 non-null     float64
 14  Es_5q3       62 non-null     float64
 15  Es_6q1       62 non-null     float64
 16  Es_6q2       62 non-null     float64
 17  final_score  62 non-null     float64
 18  pass         62 non-null     int64  
dtypes: float64(17), int64(2)
memory usage: 9.3 KB
In [103]:
pts_final_lookup.columns
Out[103]:
Index(['question', 'question_points'], dtype='object')
In [104]:
pts_final = pts_final_lookup[['question', 'question_points']].copy()
In [105]:
fig, ax = plt.subplots(figsize=(8, 6))

sns.histplot(data=final_df, x='final_score', ax=ax, bins=10)

plt.show()
In [106]:
sns.catplot(data = final_df, x='pass', kind='count')

plt.show()
In [107]:
final_df['pass'].mean()
Out[107]:
0.25806451612903225

Look at scores for each question of the question score by student¶

In [108]:
final_lf = final_df.melt( id_vars=['sid', 'final_score', 'pass']).\
           rename(columns={"variable": "question", 
                           "value": "question_score"}).\
           copy()
In [109]:
final_lf.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 992 entries, 0 to 991
Data columns (total 5 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   sid             992 non-null    int64  
 1   final_score     992 non-null    float64
 2   pass            992 non-null    int64  
 3   question        992 non-null    object 
 4   question_score  992 non-null    float64
dtypes: float64(2), int64(2), object(1)
memory usage: 38.9+ KB
In [110]:
pts_final.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 17 entries, 0 to 16
Data columns (total 2 columns):
 #   Column           Non-Null Count  Dtype 
---  ------           --------------  ----- 
 0   question         17 non-null     object
 1   question_points  17 non-null     int64 
dtypes: int64(1), object(1)
memory usage: 400.0+ bytes
In [111]:
final_comp_lf = pd.merge(final_lf, pts_final, on=['question']).copy()
In [112]:
final_comp_lf.sort_values(by=['sid'])
Out[112]:
sid final_score pass question question_score question_points
0 1 94.5 1 Es_1q1 2.0 2
124 1 94.5 1 Es_2q1 1.0 2
806 1 94.5 1 Es_5q3 3.0 3
620 1 94.5 1 Es_4q2 10.0 10
868 1 94.5 1 Es_6q1 25.0 25
... ... ... ... ... ... ...
185 106 71.0 1 Es_2q1 1.0 2
123 106 71.0 1 Es_1q2 3.0 3
61 106 71.0 1 Es_1q1 2.0 2
929 106 71.0 1 Es_6q1 25.0 25
991 106 71.0 1 Es_6q2 10.0 15

992 rows × 6 columns

In [113]:
final_comp_lf['question_pct'] = final_comp_lf['question_score'] / final_comp_lf['question_points']
In [114]:
sns.catplot(data=final_comp_lf, x='question', y='question_pct', 
            col = 'sid', kind='bar', color='slategray',
            col_wrap=1, sharey=False, sharex=False, 
            height=3, aspect=4)

plt.show()

Explore input unique counts and output data¶

Concatenate low and high participation student dataframes and join with interim_lf¶

In [115]:
high_counts_interim = pd.merge(interim_lf, sid_3high_sess_counts, on=['sid', 'sess'])
In [116]:
high_counts_interim.head()
Out[116]:
sid interim_session interim_score sess exer_uniq actv_uniq actv_grp_uniq
0 32 interim_scr2 5.5 2 7 25 9
1 44 interim_scr2 5.5 2 7 26 9
2 85 interim_scr2 6.0 2 7 24 8
3 32 interim_scr3 4.0 3 5 18 8
4 44 interim_scr3 4.0 3 5 19 9
In [117]:
low_counts_interim = pd.merge(interim_lf, sid_3low_sess_counts, on=['sid', 'sess'])
In [118]:
low_counts_interim.head()
Out[118]:
sid interim_session interim_score sess exer_uniq actv_uniq actv_grp_uniq
0 89 interim_scr3 1.5 3 4 11 7
1 57 interim_scr4 4.5 4 6 16 6
2 77 interim_scr4 4.0 4 6 16 6
3 89 interim_scr4 3.5 4 6 17 7
4 89 interim_scr5 2.5 5 2 6 5
In [119]:
var_list = uniq_vars + ['interim_score']
In [120]:
high_counts_interim.reset_index()[['sid','sess', var_list[0]]].pivot(index='sid', columns='sess', values=var_list[0])
Out[120]:
sess 2 3 4 5 6
sid
32 7 5 6 5 7
44 7 5 6 5 6
85 7 5 6 6 6

Heatmaps show participation and interim score for 3 low and 3 high participation students¶

In [121]:
fig, axs = plt.subplots(1, len(var_list), figsize=(18,6), sharex=True, sharey=True)

for ix, x in enumerate(var_list):
    sns.heatmap(data = high_counts_interim.reset_index()[['sid','sess', x]].pivot(index='sid', columns='sess', values=x),
                center=high_counts_interim.reset_index()[x].mean(),
                cmap='coolwarm', cbar=False,
                annot=True, annot_kws={'size': 25},
                ax=axs[ix])
    
    axs[ix].set_title(x)

plt.show()
In [122]:
fig, axs = plt.subplots(1, len(var_list), figsize=(18,6), sharex=True, sharey=True)

for ix, x in enumerate(var_list):
    sns.heatmap(data = low_counts_interim.reset_index()[['sid','sess', x]].pivot(index='sid', columns='sess', values=x),
                center=low_counts_interim.reset_index()[x].mean(),
                cmap='coolwarm', cbar=False,
                annot=True, annot_kws={'size': 25},
                ax=axs[ix])
    
    axs[ix].set_title(x)

plt.show()

Scatter plots show relationship between unique counts of exercises, activities, and activity groups and students by session and interim score for all students¶

In [123]:
all_counts_interim = pd.merge(interim_lf, sid_sess_counts, on=['sid', 'sess'])
In [124]:
sns.relplot(data = all_counts_interim, x='sid', y='exer_uniq', col='sess', col_wrap=3, hue='interim_score')

plt.show()
In [125]:
sns.relplot(data = all_counts_interim, x='sid', y='actv_uniq', col='sess', col_wrap=3, hue='interim_score')

plt.show()
In [126]:
sns.relplot(data = all_counts_interim, x='sid', y='actv_grp_uniq', col='sess', col_wrap=3, hue='interim_score')

plt.show()

Scatter plots show relationship between students and interim scores by unique counts of activity groups and sessions for all students¶

In [127]:
sns.relplot(data = all_counts_interim, x='sid', y='interim_score', col='actv_grp_uniq', col_wrap=3, hue='sess')

plt.show()

Evaluate student progress with final scores¶

In [128]:
final_df.head()
Out[128]:
sid Es_1q1 Es_1q2 Es_2q1 Es_2q2 Es_3q1 Es_3q2 Es_3q3 Es_3q4 Es_3q5 Es_4q1 Es_4q2 Es_5q1 Es_5q2 Es_5q3 Es_6q1 Es_6q2 final_score pass
0 1 2.0 3.0 1.0 0.5 1.0 2.0 2.0 2.0 3.0 15.0 10.0 2.0 10.0 3.0 25.0 13.0 94.5 1
1 2 2.0 3.0 2.0 0.5 1.0 2.0 0.0 2.0 3.0 15.0 2.0 0.0 5.0 1.5 5.0 0.0 44.0 0
2 4 2.0 3.0 1.0 0.5 1.0 2.0 0.0 2.0 0.0 3.0 4.0 0.0 1.5 0.0 5.0 5.0 30.0 0
3 5 2.0 3.0 2.0 1.5 1.0 2.0 2.0 2.0 3.0 3.0 2.0 1.5 9.0 1.5 2.0 1.0 38.5 0
4 7 2.0 3.0 1.0 1.5 1.0 2.0 2.0 2.0 3.0 15.0 10.0 1.0 2.5 0.0 20.0 12.0 78.0 1
In [129]:
all_counts_final = pd.merge(final_df, sid_counts, on=['sid'])
In [130]:
all_counts_final.head()
Out[130]:
sid Es_1q1 Es_1q2 Es_2q1 Es_2q2 Es_3q1 Es_3q2 Es_3q3 Es_3q4 Es_3q5 ... Es_5q2 Es_5q3 Es_6q1 Es_6q2 final_score pass sess_uniq exer_uniq actv_uniq actv_grp_uniq
0 1 2.0 3.0 1.0 0.5 1.0 2.0 2.0 2.0 3.0 ... 10.0 3.0 25.0 13.0 94.5 1 5 23 78 10
1 2 2.0 3.0 2.0 0.5 1.0 2.0 0.0 2.0 3.0 ... 5.0 1.5 5.0 0.0 44.0 0 6 30 92 10
2 4 2.0 3.0 1.0 0.5 1.0 2.0 0.0 2.0 0.0 ... 1.5 0.0 5.0 5.0 30.0 0 6 27 83 10
3 5 2.0 3.0 2.0 1.5 1.0 2.0 2.0 2.0 3.0 ... 9.0 1.5 2.0 1.0 38.5 0 6 27 83 10
4 7 2.0 3.0 1.0 1.5 1.0 2.0 2.0 2.0 3.0 ... 2.5 0.0 20.0 12.0 78.0 1 6 28 88 11

5 rows × 23 columns

In [131]:
var_list_f = uniq_vars + ['final_score']
In [132]:
all_counts_final.reset_index()[['sid'] + var_list_f]
Out[132]:
sid exer_uniq actv_uniq actv_grp_uniq final_score
0 1 23 78 10 94.5
1 2 30 92 10 44.0
2 4 27 83 10 30.0
3 5 27 83 10 38.5
4 7 28 88 11 78.0
... ... ... ... ... ...
57 101 17 54 8 32.0
58 102 25 73 10 31.5
59 103 10 34 9 18.5
60 104 11 38 8 92.0
61 106 5 19 9 71.0

62 rows × 5 columns

Scatter plots show the relationship between the number of unique exercises, activities, and activity groups and the final score¶

In [133]:
sns.relplot(data = all_counts_final, x='sid', y='exer_uniq', hue='final_score')

plt.show()
In [134]:
sns.relplot(data = all_counts_final, x='sid', y='actv_uniq', hue='final_score')

plt.show()
In [135]:
sns.relplot(data = all_counts_final, x='sid', y='actv_grp_uniq', hue='final_score')

plt.show()
In [136]:
all_counts_final_lf = all_counts_final.melt(id_vars=['sid', 'final_score'], value_vars=uniq_vars).copy()
In [137]:
sns.relplot(data = all_counts_final_lf, x='sid', y='value', col='variable', hue='final_score', 
            facet_kws={'sharey': False, 'sharex': True})

plt.show()

Explore input and output data by student¶

There is a positive relationship between the final score and final questions Es_4q2, Es_6q1, and Es_6q2. If students did well on these questions, they did well on the exam. Together these questions are worth 50% of the final score.¶

The total time spent did not appear to have an effect on whether a student passed the final. However, the number of sessions and the variety of exercises and activities in which students participated did have an effect. While participating in more sessions and in a variety of exercise and activities did not guarantee a passing score, if students did not participate in at least four sessions (with the exception of two students) and in a variety of exercises and activities, they did not pass.¶

In [138]:
inputs_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 227129 entries, 0 to 227128
Data columns (total 31 columns):
 #   Column         Non-Null Count   Dtype  
---  ------         --------------   -----  
 0   sess           227129 non-null  object 
 1   sid            227129 non-null  object 
 2   start_dt       227129 non-null  object 
 3   end_dt         227128 non-null  object 
 4   exer           227129 non-null  object 
 5   actv           227129 non-null  object 
 6   actv_grp       227129 non-null  object 
 7   total_ms       227128 non-null  float64
 8   idle_ms        226971 non-null  float64
 9   mw             227129 non-null  int64  
 10  mwc            227129 non-null  int64  
 11  mcl            227129 non-null  int64  
 12  mcr            227129 non-null  int64  
 13  mm             227129 non-null  int64  
 14  ks             227129 non-null  int64  
 15  total_ms_wdz   227129 non-null  float64
 16  idle_ms_wdz    227129 non-null  float64
 17  mw_wdz         227129 non-null  float64
 18  mwc_wdz        227129 non-null  float64
 19  mcl_wdz        227129 non-null  float64
 20  mcr_wdz        227129 non-null  float64
 21  mm_wdz         227129 non-null  float64
 22  ks_wdz         227129 non-null  float64
 23  total_ms_sqrt  227128 non-null  float64
 24  idle_ms_sqrt   226971 non-null  float64
 25  mw_sqrt        227129 non-null  float64
 26  mwc_sqrt       227129 non-null  float64
 27  mcl_sqrt       227129 non-null  float64
 28  mcr_sqrt       227129 non-null  float64
 29  mm_sqrt        227129 non-null  float64
 30  ks_sqrt        227129 non-null  float64
dtypes: float64(18), int64(6), object(7)
memory usage: 53.7+ MB
In [139]:
student_summ = inputs_sqrt_df.groupby(['sid']).\
            aggregate(total_ms_sqrt_avg = ('total_ms_sqrt', 'mean'),
                      idle_ms_sqrt_avg = ('idle_ms_sqrt', 'mean'),
                      mw_sqrt_avg = ('mw_sqrt', 'mean'),
                      mwc_sqrt_avg = ('mwc_sqrt', 'mean'),
                      mcl_sqrt_avg = ('mcl_sqrt', 'mean'),
                      mcr_sqrt_avg = ('mcr_sqrt', 'mean'),
                      mm_sqrt_avg = ('mm_sqrt', 'mean'),
                      ks_sqrt_avg = ('ks_sqrt', 'mean'),
                      sess_count = ('sess', 'nunique'),
                      exer_count = ('exer', 'nunique'),
                      act_count = ('actv_grp', 'nunique')).\
            reset_index().copy()
In [140]:
student_summ.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 113 entries, 0 to 112
Data columns (total 12 columns):
 #   Column             Non-Null Count  Dtype  
---  ------             --------------  -----  
 0   sid                113 non-null    int64  
 1   total_ms_sqrt_avg  113 non-null    float64
 2   idle_ms_sqrt_avg   113 non-null    float64
 3   mw_sqrt_avg        113 non-null    float64
 4   mwc_sqrt_avg       113 non-null    float64
 5   mcl_sqrt_avg       113 non-null    float64
 6   mcr_sqrt_avg       113 non-null    float64
 7   mm_sqrt_avg        113 non-null    float64
 8   ks_sqrt_avg        113 non-null    float64
 9   sess_count         113 non-null    int64  
 10  exer_count         113 non-null    int64  
 11  act_count          113 non-null    int64  
dtypes: float64(8), int64(4)
memory usage: 10.7 KB
In [141]:
student_sess_summ = inputs_sqrt_df.groupby(['sid', 'sess', 'actv_grp']).\
            aggregate(total_ms_sqrt_avg = ('total_ms_sqrt', 'mean'),
                      idle_ms_sqrt_avg = ('idle_ms_sqrt', 'mean'),
                      mw_sqrt_avg = ('mw_sqrt', 'mean'),
                      mwc_sqrt_avg = ('mwc_sqrt', 'mean'),
                      mcl_sqrt_avg = ('mcl_sqrt', 'mean'),
                      mcr_sqrt_avg = ('mcr_sqrt', 'mean'),
                      mm_sqrt_avg = ('mm_sqrt', 'mean'),
                      ks_sqrt_avg = ('ks_sqrt', 'mean')).\
            reset_index().copy()
In [142]:
student_sess_summ
Out[142]:
sid sess actv_grp total_ms_sqrt_avg idle_ms_sqrt_avg mw_sqrt_avg mwc_sqrt_avg mcl_sqrt_avg mcr_sqrt_avg mm_sqrt_avg ks_sqrt_avg
0 1 1 Aulaweb 60.291591 89.807296 0.446259 0.0 1.532660 0.000000 15.575328 0.000000
1 1 1 Blank 23.650093 6.663881 0.000000 0.0 1.321056 0.209706 11.315140 0.031463
2 1 1 Deeds 84.505693 145.819385 0.065204 0.0 2.423351 0.322144 22.286720 0.331025
3 1 1 Diagram 181.633786 658.822559 0.085714 0.0 4.463416 0.587310 37.397646 0.057143
4 1 1 Other 44.700700 31.988323 0.159417 0.0 1.519802 0.047670 13.975898 0.329639
... ... ... ... ... ... ... ... ... ... ... ...
4243 114 1 Diagram 158.952642 330.417026 0.000000 0.0 4.270376 0.056569 19.365146 0.080000
4244 114 1 Other 57.737868 64.730754 0.000000 0.0 1.216363 0.022688 6.911911 0.911578
4245 114 1 Properties 24.490637 13.140994 0.000000 0.0 1.242434 0.000000 5.467689 0.000000
4246 114 1 Study 74.328287 67.648816 0.000000 0.0 1.364039 0.000000 7.665700 0.133333
4247 114 1 TextEditor 67.980222 80.704891 0.000000 0.0 1.288327 0.000000 6.409345 0.812770

4248 rows × 11 columns

In [143]:
student_sess_summ.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4248 entries, 0 to 4247
Data columns (total 11 columns):
 #   Column             Non-Null Count  Dtype  
---  ------             --------------  -----  
 0   sid                4248 non-null   int64  
 1   sess               4248 non-null   int64  
 2   actv_grp           4248 non-null   object 
 3   total_ms_sqrt_avg  4248 non-null   float64
 4   idle_ms_sqrt_avg   4248 non-null   float64
 5   mw_sqrt_avg        4248 non-null   float64
 6   mwc_sqrt_avg       4248 non-null   float64
 7   mcl_sqrt_avg       4248 non-null   float64
 8   mcr_sqrt_avg       4248 non-null   float64
 9   mm_sqrt_avg        4248 non-null   float64
 10  ks_sqrt_avg        4248 non-null   float64
dtypes: float64(8), int64(2), object(1)
memory usage: 365.2+ KB

Inputs and Intermediate Scores¶

In [144]:
interim_lf.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 575 entries, 0 to 574
Data columns (total 4 columns):
 #   Column           Non-Null Count  Dtype  
---  ------           --------------  -----  
 0   sid              575 non-null    int64  
 1   interim_session  575 non-null    object 
 2   interim_score    575 non-null    float64
 3   sess             575 non-null    int64  
dtypes: float64(1), int64(2), object(1)
memory usage: 18.1+ KB
In [145]:
interim_lf['sid'] = interim_lf['sid'].astype('int')
interim_lf['sess'] = interim_lf['sess'].astype('int')
In [146]:
by_student_interim_init = pd.merge(interim_lf.drop(columns=['interim_session'].copy()), student_sess_summ, 
                                   on=['sid','sess'], how='left')
In [147]:
by_student_interim_init.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 3779 entries, 0 to 3778
Data columns (total 12 columns):
 #   Column             Non-Null Count  Dtype  
---  ------             --------------  -----  
 0   sid                3779 non-null   int64  
 1   interim_score      3779 non-null   float64
 2   sess               3779 non-null   int64  
 3   actv_grp           3642 non-null   object 
 4   total_ms_sqrt_avg  3642 non-null   float64
 5   idle_ms_sqrt_avg   3642 non-null   float64
 6   mw_sqrt_avg        3642 non-null   float64
 7   mwc_sqrt_avg       3642 non-null   float64
 8   mcl_sqrt_avg       3642 non-null   float64
 9   mcr_sqrt_avg       3642 non-null   float64
 10  mm_sqrt_avg        3642 non-null   float64
 11  ks_sqrt_avg        3642 non-null   float64
dtypes: float64(9), int64(2), object(1)
memory usage: 383.8+ KB
In [148]:
by_student_interim_init
Out[148]:
sid interim_score sess actv_grp total_ms_sqrt_avg idle_ms_sqrt_avg mw_sqrt_avg mwc_sqrt_avg mcl_sqrt_avg mcr_sqrt_avg mm_sqrt_avg ks_sqrt_avg
0 1 5.0 2 Aulaweb 19.023607 5.364728 0.000000 0.0 1.660312 0.000000 9.816710 0.000000
1 1 5.0 2 Blank 13.122374 4.530336 0.039229 0.0 1.335669 0.124054 9.710477 0.000000
2 1 5.0 2 Deeds 94.955802 196.928061 0.168232 0.0 2.899294 0.408978 23.214648 0.366994
3 1 5.0 2 Diagram 102.537769 223.944413 0.000000 0.0 2.703978 0.036262 23.804276 0.000000
4 1 5.0 2 Other 40.282864 13.255219 0.420092 0.0 1.743471 0.000000 13.599805 0.161676
... ... ... ... ... ... ... ... ... ... ... ... ...
3774 111 0.0 6 NaN NaN NaN NaN NaN NaN NaN NaN NaN
3775 112 0.0 6 NaN NaN NaN NaN NaN NaN NaN NaN NaN
3776 113 0.0 6 NaN NaN NaN NaN NaN NaN NaN NaN NaN
3777 114 0.0 6 NaN NaN NaN NaN NaN NaN NaN NaN NaN
3778 115 0.0 6 NaN NaN NaN NaN NaN NaN NaN NaN NaN

3779 rows × 12 columns

In [149]:
by_student_interim_df = by_student_interim_init.copy()
In [150]:
by_student_interim_lf = by_student_interim_df.melt(id_vars=['sid', 'sess', 'actv_grp', 'interim_score'])
In [151]:
by_student_interim_lf
Out[151]:
sid sess actv_grp interim_score variable value
0 1 2 Aulaweb 5.0 total_ms_sqrt_avg 19.023607
1 1 2 Blank 5.0 total_ms_sqrt_avg 13.122374
2 1 2 Deeds 5.0 total_ms_sqrt_avg 94.955802
3 1 2 Diagram 5.0 total_ms_sqrt_avg 102.537769
4 1 2 Other 5.0 total_ms_sqrt_avg 40.282864
... ... ... ... ... ... ...
30227 111 6 NaN 0.0 ks_sqrt_avg NaN
30228 112 6 NaN 0.0 ks_sqrt_avg NaN
30229 113 6 NaN 0.0 ks_sqrt_avg NaN
30230 114 6 NaN 0.0 ks_sqrt_avg NaN
30231 115 6 NaN 0.0 ks_sqrt_avg NaN

30232 rows × 6 columns

In [152]:
sns.catplot(data=by_student_interim_lf, x = 'actv_grp', y = 'interim_score', 
            kind = 'box', palette='deep',
            hue = 'actv_grp', height=3, aspect=4, 
            sharey=False, sharex=False)

plt.show()
In [153]:
sns.catplot(data=by_student_interim_lf, x = 'variable', y = 'interim_score', 
            col = 'sess', col_wrap = 2, kind = 'box', palette='deep',
            hue = 'actv_grp', height=3, aspect=4, 
            sharey=False, sharex=False)

plt.show()
In [154]:
sns.lmplot(data = by_student_interim_lf.loc[by_student_interim_lf.sess == 2], x='value', y='interim_score', 
           col='variable', row='actv_grp', hue='actv_grp', palette='deep',
           aspect=2, facet_kws=dict(sharex=False, sharey=False))

plt.show()
In [155]:
by_student_interim_lf.variable.unique()
Out[155]:
array(['total_ms_sqrt_avg', 'idle_ms_sqrt_avg', 'mw_sqrt_avg',
       'mwc_sqrt_avg', 'mcl_sqrt_avg', 'mcr_sqrt_avg', 'mm_sqrt_avg',
       'ks_sqrt_avg'], dtype=object)
In [156]:
sns.lmplot(data = by_student_interim_lf.loc[(by_student_interim_lf['variable'] == 'total_ms_sqrt_avg') & (by_student_interim_lf['sess'] == 2)], 
           x='value', y='interim_score', 
           col='actv_grp', col_wrap = 2, hue='actv_grp', palette='deep',
           aspect=2, facet_kws=dict(sharex=False, sharey=False))

plt.show()
In [157]:
sns.lmplot(data = by_student_interim_lf.loc[(by_student_interim_lf['variable'] == 'idle_ms_sqrt_avg') & (by_student_interim_lf['sess'] == 2)], 
           x='value', y='interim_score', 
           col='actv_grp', col_wrap = 2, hue='actv_grp', palette='deep',
           aspect=2, facet_kws=dict(sharex=False, sharey=False))

plt.show()
In [158]:
sns.lmplot(data = by_student_interim_lf.loc[(by_student_interim_lf['variable'] == 'mcr_sqrt_avg') & (by_student_interim_lf['sess'] == 2)], 
           x='value', y='interim_score', 
           col='actv_grp', col_wrap = 2, hue='actv_grp', palette='deep',
           aspect=2, facet_kws=dict(sharex=False, sharey=False))

plt.show()
In [159]:
sns.lmplot(data = by_student_interim_lf.loc[(by_student_interim_lf['variable'] == 'ks_sqrt_avg') & (by_student_interim_lf['sess'] == 2)], 
           x='value', y='interim_score', 
           col='actv_grp', col_wrap = 2, hue='actv_grp', palette='deep',
           aspect=2, facet_kws=dict(sharex=False, sharey=False))

plt.show()

Inputs and Final Score¶

In [160]:
final_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 62 entries, 0 to 61
Data columns (total 19 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   sid          62 non-null     int64  
 1   Es_1q1       62 non-null     float64
 2   Es_1q2       62 non-null     float64
 3   Es_2q1       62 non-null     float64
 4   Es_2q2       62 non-null     float64
 5   Es_3q1       62 non-null     float64
 6   Es_3q2       62 non-null     float64
 7   Es_3q3       62 non-null     float64
 8   Es_3q4       62 non-null     float64
 9   Es_3q5       62 non-null     float64
 10  Es_4q1       62 non-null     float64
 11  Es_4q2       62 non-null     float64
 12  Es_5q1       62 non-null     float64
 13  Es_5q2       62 non-null     float64
 14  Es_5q3       62 non-null     float64
 15  Es_6q1       62 non-null     float64
 16  Es_6q2       62 non-null     float64
 17  final_score  62 non-null     float64
 18  pass         62 non-null     int64  
dtypes: float64(17), int64(2)
memory usage: 9.3 KB
In [161]:
final_df.columns
Out[161]:
Index(['sid', 'Es_1q1', 'Es_1q2', 'Es_2q1', 'Es_2q2', 'Es_3q1', 'Es_3q2',
       'Es_3q3', 'Es_3q4', 'Es_3q5', 'Es_4q1', 'Es_4q2', 'Es_5q1', 'Es_5q2',
       'Es_5q3', 'Es_6q1', 'Es_6q2', 'final_score', 'pass'],
      dtype='object')
In [162]:
by_student_final_init = pd.merge(student_summ, final_df, on='sid', how='left')
In [163]:
by_student_final_init.rename(columns={"Es_1_1": "final_Es_1_1", "Es_1_2": "final_Es_1_2", "Es_2_1": "final_Es_2_1", 
                              "Es_2_2": "final_Es_2_2", "Es_3_1": "final_Es_3_1", "Es_3_2": "final_Es_3_2", 
                              "Es_3_3": "final_Es_3_3", "Es_3_4": "final_Es_3_4", "Es_3_5": "final_Es_3_5", 
                              "Es_4_1": "final_Es_4_1", "Es_4_2": "final_Es_4_2", "Es_5_1": "final_Es_5_1", 
                              "Es_5_2": "final_Es_5_2", "Es_5_3": "final_Es_5_3", "Es_6_1": "final_Es_6_1", 
                              "Es_6_2": "final_Es_6_2"}, inplace=True)
In [164]:
by_student_final_init.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 113 entries, 0 to 112
Data columns (total 30 columns):
 #   Column             Non-Null Count  Dtype  
---  ------             --------------  -----  
 0   sid                113 non-null    int64  
 1   total_ms_sqrt_avg  113 non-null    float64
 2   idle_ms_sqrt_avg   113 non-null    float64
 3   mw_sqrt_avg        113 non-null    float64
 4   mwc_sqrt_avg       113 non-null    float64
 5   mcl_sqrt_avg       113 non-null    float64
 6   mcr_sqrt_avg       113 non-null    float64
 7   mm_sqrt_avg        113 non-null    float64
 8   ks_sqrt_avg        113 non-null    float64
 9   sess_count         113 non-null    int64  
 10  exer_count         113 non-null    int64  
 11  act_count          113 non-null    int64  
 12  Es_1q1             62 non-null     float64
 13  Es_1q2             62 non-null     float64
 14  Es_2q1             62 non-null     float64
 15  Es_2q2             62 non-null     float64
 16  Es_3q1             62 non-null     float64
 17  Es_3q2             62 non-null     float64
 18  Es_3q3             62 non-null     float64
 19  Es_3q4             62 non-null     float64
 20  Es_3q5             62 non-null     float64
 21  Es_4q1             62 non-null     float64
 22  Es_4q2             62 non-null     float64
 23  Es_5q1             62 non-null     float64
 24  Es_5q2             62 non-null     float64
 25  Es_5q3             62 non-null     float64
 26  Es_6q1             62 non-null     float64
 27  Es_6q2             62 non-null     float64
 28  final_score        62 non-null     float64
 29  pass               62 non-null     float64
dtypes: float64(26), int64(4)
memory usage: 27.4 KB
In [165]:
by_student_final_df = by_student_final_init.copy()
In [166]:
sns.relplot(data = by_student_final_df, x='total_ms_sqrt_avg', y='final_score', kind='scatter')

plt.show()
In [167]:
by_student_final_df.columns
Out[167]:
Index(['sid', 'total_ms_sqrt_avg', 'idle_ms_sqrt_avg', 'mw_sqrt_avg',
       'mwc_sqrt_avg', 'mcl_sqrt_avg', 'mcr_sqrt_avg', 'mm_sqrt_avg',
       'ks_sqrt_avg', 'sess_count', 'exer_count', 'act_count', 'Es_1q1',
       'Es_1q2', 'Es_2q1', 'Es_2q2', 'Es_3q1', 'Es_3q2', 'Es_3q3', 'Es_3q4',
       'Es_3q5', 'Es_4q1', 'Es_4q2', 'Es_5q1', 'Es_5q2', 'Es_5q3', 'Es_6q1',
       'Es_6q2', 'final_score', 'pass'],
      dtype='object')
In [168]:
by_student_final_lf = by_student_final_df.melt(id_vars=['sid', 'final_score', 'pass'])
In [169]:
by_student_final_lf.isna().sum()
Out[169]:
sid               0
final_score    1377
pass           1377
variable          0
value           816
dtype: int64
In [170]:
by_student_final_lf.variable.unique()
Out[170]:
array(['total_ms_sqrt_avg', 'idle_ms_sqrt_avg', 'mw_sqrt_avg',
       'mwc_sqrt_avg', 'mcl_sqrt_avg', 'mcr_sqrt_avg', 'mm_sqrt_avg',
       'ks_sqrt_avg', 'sess_count', 'exer_count', 'act_count', 'Es_1q1',
       'Es_1q2', 'Es_2q1', 'Es_2q2', 'Es_3q1', 'Es_3q2', 'Es_3q3',
       'Es_3q4', 'Es_3q5', 'Es_4q1', 'Es_4q2', 'Es_5q1', 'Es_5q2',
       'Es_5q3', 'Es_6q1', 'Es_6q2'], dtype=object)
In [171]:
sns.relplot(data = by_student_final_lf, x='value', y='final_score', 
            col='variable', col_wrap=2, 
            kind='scatter', facet_kws={'sharex': False})

plt.show()
In [172]:
pts_final
Out[172]:
question question_points
0 Es_1q1 2
1 Es_1q2 3
2 Es_2q1 2
3 Es_2q2 3
4 Es_3q1 1
5 Es_3q2 2
6 Es_3q3 2
7 Es_3q4 2
8 Es_3q5 3
9 Es_4q1 15
10 Es_4q2 10
11 Es_5q1 2
12 Es_5q2 10
13 Es_5q3 3
14 Es_6q1 25
15 Es_6q2 15
16 final_score 100
In [173]:
by_student_final_df
Out[173]:
sid total_ms_sqrt_avg idle_ms_sqrt_avg mw_sqrt_avg mwc_sqrt_avg mcl_sqrt_avg mcr_sqrt_avg mm_sqrt_avg ks_sqrt_avg sess_count ... Es_3q5 Es_4q1 Es_4q2 Es_5q1 Es_5q2 Es_5q3 Es_6q1 Es_6q2 final_score pass
0 1 73.788436 154.779253 0.438855 0.000000 2.154162 0.100167 18.288536 0.575429 5 ... 3.0 15.0 10.0 2.0 10.0 3.0 25.0 13.0 94.5 1.0
1 2 89.933800 302.859102 0.492373 0.000000 2.087846 0.149846 14.232934 0.924389 6 ... 3.0 15.0 2.0 0.0 5.0 1.5 5.0 0.0 44.0 0.0
2 3 67.607202 177.972725 0.200320 0.000000 2.010457 0.146013 13.700542 0.648768 4 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
3 4 62.942091 158.045946 0.389018 0.003008 1.661840 0.055576 9.571955 0.721963 6 ... 0.0 3.0 4.0 0.0 1.5 0.0 5.0 5.0 30.0 0.0
4 5 70.455967 149.440560 0.512454 0.000000 1.993788 0.150808 19.069058 0.673137 6 ... 3.0 3.0 2.0 1.5 9.0 1.5 2.0 1.0 38.5 0.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
108 110 69.048451 157.689999 0.371386 0.000000 2.058844 0.106372 14.245068 0.547387 1 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
109 111 72.166219 96.645577 0.000000 0.000000 2.087146 0.213448 13.376083 0.407551 1 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
110 112 82.276579 202.644242 0.234535 0.003961 1.934829 0.106930 17.046930 0.468303 1 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
111 113 81.240771 177.109727 0.000000 0.000000 1.940302 0.183114 12.548179 0.373359 1 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
112 114 65.691443 107.541299 0.000000 0.000000 1.531306 0.022377 7.922776 0.487939 1 ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN

113 rows × 30 columns

In [174]:
sns.lmplot(data = by_student_final_df, x='Es_4q2', y='final_score')

plt.show()
In [175]:
sns.lmplot(data = by_student_final_df, x='Es_6q1', y='final_score')

plt.show()
In [176]:
sns.lmplot(data = by_student_final_df, x='Es_6q2', y='final_score')

plt.show()

The box plots below show that participation in the session, exercises, and activities influences final grades¶

In [177]:
sns.catplot(data=by_student_final_df, 
            y='total_ms_sqrt_avg', col='pass', 
            kind='box', 
            color='slategray',
            sharex = False, height=2, aspect=2)

plt.show()
In [178]:
sns.catplot(data=by_student_final_df, 
            y='sess_count', col='pass', 
            kind='box', 
            color='slategray',
            sharex = False, height=2, aspect=2)

plt.show()
In [179]:
sns.catplot(data=by_student_final_df, 
            y='exer_count', col='pass', 
            kind='box',
            color='slategray',
            sharex = False, height=2, aspect=2)

plt.show()
In [180]:
sns.catplot(data=by_student_final_df, 
            y='act_count', col='pass', 
            kind='box',
            color='slategray',
            sharex = False, height=2, aspect=2)

plt.show()
In [ ]: